AI maturity in the manufacturing sector
Your industrial AI trajectory, measured site by site and turned into a costed action plan.
10 themes, a 5-level scale. And the action that moves each level to the next.
The framework’s 10 themes, already written from L1 to L5. One company, one business unit, or 300 at once.
AI maturity in the manufacturing sector
10 themes, 5-level scale.
Nordhavn Industries
53 / 100
They measure their maturity with Datamensio
An example
This could be your situation.
Take one company as an example: three sites, three spreadsheets, no shared answer.
Nobody can consolidate.
Nordhavn Industries, 2,400 people in Hamburg, Lyon and Porto. A client asks where the group stands. Each site answers in its own spreadsheet, with its own scales.
Three weeks, a single base.
One Industrial AI maturity framework (Datamensio framework, CMMI scale) assessment launched across all three sites at once, from the managers’ interview notes. The framework was already written, its 10 themes and levels L1 to L5 too.
Two costs avoided before being committed.
A score of 53 out of 100, with the gap concentrated on three themes. The AI companion spotted that two actions duplicated those of another audit. The committee report took one sentence to request.
What it saved them
- 3sites measured on the same base, instead of three questionnaires to reconcile
- 2duplicate actions caught before the spend
- 1committee report, with no manual rework
These figures are an example. They could be yours.
The standard imposes processes. Datamensio says where you stand.
01
The framework is already written
Themes, questions and levels L1 to L5, all written. You do not start from an empty spreadsheet.
02
The score lands the same day
Online, by self-assessment link or in interview. Theme by theme, comparable over time.
03
The gap becomes a costed plan
Every step up carries its action. The AI prioritises on expected effect, not on the order of the standard.
04
Progress can be demonstrated
Campaign after campaign, against your target and against your own past. That is what your board asks for.
The maturity scale
One level, the next, and the action that links the two.
It is this mechanism (a level, a higher level, and the action linking the two) that turns an observation into a trajectory.
How are AI models monitored once deployed in production on the lines?
- N1
No monitoring. Once deployed, the model runs without its performance being tracked. Deviations are discovered by line teams.
- N2
Monitoring exists on a few models, on the initiative of whoever built them. Indicators and frequency vary from one use case to another.
- N3
Every model in production has defined performance indicators and an identified owner. Deviations are reviewed at a periodic meeting.
- N4
Drift is detected automatically, with alert thresholds and a documented retraining procedure. Versions and associated decisions are traced.
- N5
Thresholds and procedures are revised based on operational feedback and changes to the product range. Lessons learned are shared with other sites and folded into the group standard.
Action to move from L2 to L3
Draw up the list of models actually in operation, name a site side owner for each, define two or three performance indicators, and add their review to the agenda of the plant’s monthly performance meeting.
« With Datamensio, we meet our objectives far more efficiently. The ERDF inspection services and our supervising ministry particularly appreciated an approach that gives them reliable data. »
Director, CCI 94CCI Île-de-France
« We believe this is the most suitable solution to scale our transformation project and measure impact according to our needs. »
Maja SucekChief Operating Officer, Interreg Danube
Rarely on its own
Frameworks combine. Put several together to cover your business, or have the AI write yours.
Take your first measurement
What this framework covers
AI maturity in the manufacturing sector is measured across the whole chain, from shop floor data to decisions taken in production. It covers the availability and quality of machine, sensor, MES and CMMS data, the ability to build and validate models, their deployment in constrained environments (cycle time, OT network, line stoppages), adoption by production teams, and the governance that frames all of it. A single use case that works on one site says nothing about this maturity.
Steering almost always runs into the same issues. How many proofs of concept have made it into continuous operation, and how many stayed on a data scientist’s laptop? Who decides that a quality control model can downgrade a part without human validation, and on what criteria? What happens when the product reference changes and the model drifts: is there monitoring, an alert, a planned retraining? These questions are not settled at group level. They are settled site by site, line by line.
A common confusion keeps coming back: industrial AI is often assumed to be a natural extension of Industry 4.0 and equipment connectivity projects. Connectivity produces data, it does not produce AI maturity. Many plants hold rich historical data and run very few models in production, for lack of data ownership, engineering resources, or a validation process accepted by quality and production. Conversely, lightly instrumented sites get results because they scoped a narrow use case and industrialised it.
A binary state of play (the site does AI, or it does not) does not help decision making. The maturity assessment places each theme on a progressive scale, describes the observable practice at each level, and names the action that moves it up a level. This is what allows two business units to be compared, a differentiated target to be set by site, and the effort to be costed before it is committed. The services catalogue matches a solution to every item in the action plan, with its cost, timeline and impact on the score.
The framework is ready to use and belongs to you. Datamensio’s AI adjusts the themes, questions and levels to your industrial organisation, or builds a variant from your own documents: master plan, group standards, committee minutes. You can also start from a blank template and design your own grid in Maps.
Reference standard: Industrial AI maturity framework (Datamensio framework, CMMI scale)
The themes assessed
Strategy and use case portfolio
Existence of an AI trajectory tied to industrial objectives, use case selection criteria, arbitration between sites, tracking of value generated after go live.
Production data
Equipment coverage, historisation, quality and completeness of time series, product and routing master data, contextualisation of machine data, accessibility for project teams.
Technical foundation and OT integration
Processing platform, edge and central execution capabilities, connection to MES, CMMS, SCADA and PLCs, cycle time and network availability constraints.
Model design and validation
Scoping method, representative test sets, performance criteria accepted by production and quality, documentation, reproducibility of training runs.
Deployment and monitoring
Moving from experimentation to continuous production, drift monitoring, retraining procedure, version management, fallback procedure when a model is unavailable.
Adoption by production teams
Integration into workstations and line routines, readability of recommendations for operators, handling of false positives, feedback loop to technical teams.
Skills and organisation
Distribution of roles between site, group and providers, presence of a champion per plant, upskilling plan for automation engineers and process technicians.
Governance, ethics and regulatory framework
Validation bodies, classification of uses by criticality, traceability of automated decisions, consideration of the EU AI Act and related documentation obligations.
Cybersecurity and industrial data control
Segmentation between IT and OT networks, control of outbound flows, ownership of data shared with equipment suppliers, protection of models and process know how.
Results measurement and industrialisation
Performance indicators tied to each use case, before and after comparison, replication from one site to another, capitalising on failures.
A short version of the framework is available for the online self-assessment.
Frequently asked questions
Does this framework lead to a certification?
No. There is no body that certifies a plant’s AI maturity. The assessment measures your practices on a progressive scale and produces an action plan. It is a decision and steering tool, not a label.
How is this different from a compliance audit?
An audit checks for the presence of requirements and concludes with a gap or a pass. The maturity assessment places each theme on five levels and names the action that moves it up a level. On a transformation topic like industrial AI, what matters is the trajectory, not the verdict.
How long does the assessment take?
The short version takes 20 to 30 minutes to complete for a manager who knows the scope. The full version, run collaboratively with data, production, quality and automation teams, generally spans one to two weeks, with most of the time spent on data collection.
Can the framework be adapted to our operations?
Yes. The themes, questions and levels can be modified, and the AI generates a variant adapted to your sector, process or discrete manufacturing, based on your internal documents. You can also add your own themes or start from a blank template.
How do we compare several plants against each other?
The assessment is rolled out across as many sites as needed, using the same grid. The benchmark positions each business unit against the others and against its own previous results. A cross site roadmap consolidates action plans from several sites into pooled actions.
Do respondents need data science expertise?
The questions focus on practices, organisation and operation, not on algorithms. An industrial manager or project lead can answer them. Technical points can be assigned to a data or automation contact in collaborative mode.
Is the EU AI Act taken into account?
Yes, within the governance theme: classification of uses by criticality, traceability of automated decisions, expected documentation. The assessment measures your level of readiness on these points, it does not replace a legal analysis.
Where is the data hosted?
In France, with OVH, backed up with Scaleway. No transfer outside the European Union. The AI models used within the platform can be selected, including from European solutions.